After bug 558287 landed browser_searching.js, it produces numerous failures on SeaMonkey. I suspect that adding the prefs it added in firefox.js on our side will fix those as well.
Preliminary testing doesn't look that good with just those prefs added. Someone needs to take a deeper look into this, either me or preferably someone else.
Do you refer to the *test* file that was added? This bug report is quite vague, I thought it referred to an actual functionality breakage.
yes, browser_searching.js is a test file and it produces about 80 or so mochitest-browser-chrome test failures.
Created attachment 459523 [details] [diff] [review] fix the prefs (not the test) [Checkin: Comment 6] This patch just fixes the prefs. It doesn't really help the test, I guess it has clever design that hardwires them to some value anyhow and uses local mochi.test URLs instead of querying the real AMO service anyhow. Still, we should set those in our build in any case, so I guess I should better attach them and get them reviewed, right? ;-)
Comment on attachment 459523 [details] [diff] [review] fix the prefs (not the test) [Checkin: Comment 6] seems to work: https://services.addons.mozilla.org/en-US/firefox/api/1/search/chatzilla/all/2/windows/2.0 and https://services.addons.mozilla.org/en-US/seamonkey/api/1/search/chatzilla/all/2/windows/2.0 return the same data... (granted I used a pretty contrived search term and an app version we won't see on our nightlies, but still.
Comment on attachment 459523 [details] [diff] [review] fix the prefs (not the test) [Checkin: Comment 6] http://hg.mozilla.org/comm-central/rev/91d9bc2eeaa3 Lets see where this takes things to start with
Looks like bug 566593 actually fixed that, but I'll mark this as mine for the patch that was checked in here.